A Hybrid Context-aware Service Platform Based on Stochastic and Rule-Description Approaches

A key challenge of ubiquitous computing environments, which feature massively distributed systems containing a large number of contexts and services, is how to manage the relations between context and service. Some researchers use rules to describe these relations, while other researchers propose stochastic methods to model them. However, both approaches have serious problems cannot be solved by themselves. Therefore, in this paper, we present a hybrid context-aware service platform, which integrate the strong points of both stochastic and rule-description approaches. With our original stochastic model, most relations can be automatically learned from the recorded history of context and service; for those cannot be processed via the stochastic model, we explicitly define them as rules. We are developing a prototype based on this platform, which can recommend safe and appropriate services adapting to dynamic contexts in an indoor environment.